Academic tracking is a common feature of school organization, but it produces inequalities in student outcomes. High-stakes testing policies offer new opportunities for assessing students' progress, but the instruments raise concerns about relying on such narrow measures of learning. This study utilizes a longitudinal data set that follows one cohort of North Carolina school students from public high school through the state university system to examine the difference in outcomes associated with academic track. Results show that upper track students do better in college even when controlling for achievement on high-stakes tests and that such tests are a stronger predictor of college success for upper track students than they are for lower track students. Interviews suggest that these differences can be attributed to different methods of instruction in each track.